COGNITIVE INTERFERENCE ACROSS THE LIFESPAN

Abstract Mental processes that facilitate goal-directed behavior can be negatively affected by age-related biological and physiological changes. Such cognitive changes impact daily activities like driving and using the computer, which can in turn influence social relationships and capacity to work. Previous research has shown that older adults are more affected by interference than young adults. It is critical to understand how cognition changes across the adult lifespan, as cognition during middle age may be predictive of cognitive decline in older age. We evaluated performance on the MultiSource Interference Task (MSIT). The MSIT draws on Stroop, Flanker, and Simon-type tasks; it taxes interference resolution and typically results in interference-related slowing. We studied 60 individuals (32 adults 18–39 years old, 17 adults 40–59 years old, and 11 adults 60–99 years old). As expected, age had a significant effect on accuracy interference (difference in errors between interference and control trials). Surprisingly, however, young adults demonstrated the highest (worst) accuracy interference cost, while middle-aged and older adults had similar interference cost. Replicating previous findings, age group did not have an effect on reaction time interference cost (slowing of response between interference and control trials). It is notable that, in this study, older adults did not fare worse than middle-aged adults. These surprising findings challenge long-held theories that age negatively impacts interference resolution. Understanding differences (and similarities) between cognition in middle-age and older age will be critical for promoting healthy cognition throughout the lifespan, which benefits daily activities and quality of life.

Using cross-validated regularized regression, we identified dozens of factors acting as CR proxies (e.g., leisure reading as protective and screen usage as a risk factor during adolescence). We then use a within-family design to examine the moderation effect over and above genetic and environmental covariates. For example, adolescents who read more tend to show better cognitive performance given their gray matter volume (b = .093,95%CI = [0.035,0.152]). This study aims to identify factors that could be targeted with scalable prevention and intervention efforts earlier in life to maximize cognitive functioning.

NEW EVIDENCE OF HEALTHIER AGING. POSITIVE COHORT EFFECT ON COGNITIVE DECLINE.
Fernando Massa, 1 Alejandra Marroig, 2 Graciela Muniz Terrera, 3 and Scott Hofer 4 , 1. Universidad de la República,Montevideo,Montevideo,Uruguay,2. Universidad de la Republica Uruguay,Montevideo,Montevideo,Uruguay,3. Ohio University,Athens,Ohio,United States,4. University of Victoria,Victoria,British Columbia,Canada Cross sectional studies have shown cohort effects in cognition, limited research exists about cohort effects on cognitive trajectories. Indeed, most longitudinal research conducted to study aging-related cognitive change focus on the association between risk factors and mean change in cognition, considering individual differences too, but longitudinal norms of cognitive function are less studied. In this study, we aim to test whether cohort effects exist across the distribution of verbal fluency trajectories, that is, whether cohort effects vary across different trajectory quantiles. With this purpose, we estimated norms using data from 9 waves of the English Longitudinal Study of Aging (ELSA). We considered the individuals born in the 1920s, 1930s, and 1940s to assess cohort effects. The methodological framework consisted of quantile mixed models where the effect of age was adjusted using splines. To test for possible cohort effects across the 5th, 50th and 90th quartiles, the coefficients associated with the splines varied among cohorts. Our results suggest that cognitive decline is less pronounced for individuals born in more recent decades (p < 0.001), supporting our hypothesis of cohort effects. Moreover, these results are consistent across quantiles (p-value < 0.001). Additionally, we found that quantiles of verbal fluency at a certain age is higher in participants from more recent cohorts compared to those in older cohorts. Our findings contribute to a better understanding of cognitive decline in older adults, demonstrating population changes over time at different levels of changes in verbal fluency.

EDUCATION AND COGNITIVE ABILITY: EXPLORING DIRECT AND INDIRECT EFFECTS IN THE UNITED STATES AND MEXICO Joseph Saenz, and Amina Khan, University of Southern California, Los Angeles, California, United States
Education positively relates with cognition, which may be explained by enhanced cognitive reserve. However, education may also impact cognition indirectly by improving health, health behaviors, and life-course socioeconomic status (SES). This analysis explores the associations between education and cognition in the US and Mexico and quantifies the extent to which associations are direct versus indirect through health and SES. We use data from two studies: the MexCog in Mexico (n=2,042) and US Harmonized Cognitive Assessment Protocol (HCAP, n=3,267). Cognitive domains included Memory, Executive Function, Language, Visuospatial, and Orientation. Karlson-Holm-Breen (KHB) methods were used in linear regression models to quantify how much of the associations between years of education and cognitive domains were direct versus indirect through chronic conditions, income, wealth, smoking, and exercise. In regression models, years of education related positively with all cognitive domains in both studies, even when controlling for health and SES. KHB mediation analyses suggested that most of the education-cognition association was direct. In MexCog, estimates of the percent of the educationcognition association that was indirect through health and SES ranged from 4.17% (Memory) to 5.15% (Executive Function). In HCAP, indirect effects ranged from 8.95% (Orientation) to 12.15% (Language). Education was associated with better cognitive abilities in the US and Mexico regardless of cognitive domain or adjustment for late-life health and SES. Results suggested that education primarily related with cognition directly and that effects of education on cognitive abilities are not eliminated by reducing educational disparities in the late-life health and SES factors we analyzed.

COGNITIVE INTERFERENCE ACROSS THE LIFESPAN
Rachel Cole, Arturo Espinoza, Brooke Yeager, and Nandakumar Narayanan, University of Iowa, Iowa City, Iowa, United States Mental processes that facilitate goal-directed behavior can be negatively affected by age-related biological and physiological changes. Such cognitive changes impact daily activities like driving and using the computer, which can in turn influence social relationships and capacity to work. Previous research has shown that older adults are more affected by interference than young adults. It is critical to understand how cognition changes across the adult lifespan, as cognition during middle age may be predictive of cognitive decline in older age. We evaluated performance on the MultiSource Interference Task (MSIT). The MSIT draws on Stroop, Flanker, and Simon-type tasks; it taxes interference resolution and typically results in interference-related slowing. We studied 60 individuals (32 adults 18-39 years old, 17 adults 40-59 years old, and 11 adults 60-99 years old). As expected, age had a significant effect on accuracy interference (difference in errors between interference and control trials). Surprisingly, however, young adults demonstrated the highest (worst) accuracy interference cost, while middle-aged and older adults had similar interference cost. Replicating previous findings, age group did not have an effect on reaction time interference cost (slowing of response between interference and control trials). It is notable that, in this study, older adults did not fare worse than middleaged adults. These surprising findings challenge long-held theories that age negatively impacts interference resolution. Understanding differences (and similarities) between cognition in middle-age and older age will be critical for promoting healthy cognition throughout the lifespan, which benefits daily activities and quality of life.

CHARACTERIZING PERFORMANCE VARIABILITY ACROSS COGNITIVE TASKS IN OLDER ADULTS WITH AND WITHOUT ADHD
Laura M. Klepacz, 1 Eric Cerino, 1 and Almar A. L. Kok 2 ,

Northern Arizona University, Flagstaff, Arizona, United States, 2. Amsterdam University Medical Center, Amsterdam, Noord-Holland, Netherlands
Most research on cognitive performance among individuals with Attention Deficit/Hyperactivity Disorder (ADHD) focuses on younger persons and on cognitive variability within speeded response-time tasks. Dispersion, i.e., variability across a range of cognitive domains, is emerging as a promising indicator of age-related and pathological cognitive impairment. However, there has yet to be an evaluation of differences in dispersion among older adults with and without ADHD. We address this gap by assessing associations of age and ADHD status with dispersion. We hypothesize older adults will exhibit greater dispersion than comparatively younger adults and explore whether individuals with ADHD exhibit greater dispersion than individuals without ADHD. In a sample of 231 adults from the Longitudinal Aging Study Amsterdam (Average age=71.64 years, SD=7.7, 59% female), 23 individuals met DSM-IV criteria for ADHD and 208 were classified as neurotypically functioning. Participants completed 13 tasks spanning domains of attention, fluency, memory, processing speed, and reasoning. Dispersion across the tasks was calculated as an intraindividual standard deviation. We regressed dispersion on age and ADHD status and adjusted for sex. Older age was significantly associated with greater dispersion (Est =0.06, SE=0.03, p=0.01). However, dispersion profiles did not vary as a function of ADHD status (Est.=-0.25, SE=0.67, p>05). Preliminary results suggest that dispersion across cognitive tasks may not be a sensitive marker of ADHD in older adults, although statistical power to detect differences was relatively low in the current study. As expected, age was a significant predictor of increased dispersion, consistent with accounts of age-related changes in neurological integrity.

ASSOCIATION OF COGNITIVE STATUS AND CONSUMPTION OF UNPROCESSED AND ULTRA-PROCESSED FOODS IN BRAZILIAN OLDER ADULTS
Graziele Silva, 1 Carol Freiria, 2 Tábatta Brito, 3 Flávia Arbex Silva Borim, 2 and Ligiana P. Corona 4 , 1. UNICAMP,Limeira,Sao Paulo,Brazil,2. Universidade Estadual de Campinas,Campinas,Sao Paulo,Brazil,3. Universidade Federal de Alfenas,Alfenas,Sao Paulo,Brazil,4. Universidade Estadual de Campinas/ Faculdade de Ciências Aplicadas,Limeira,Sao Paulo,Brazil The decrease in consumption of fresh or minimally processed foods and the increase in ultra-processed foods are being observed in the diet of older adults and these changes may lead to worsening health status and cognition. We aimed to evaluate the association between cognitive status and food consumption according to the level of processing in Brazilian older adults. Cross-sectional study, with a sample of 585 older adults (≥60 years). Cognition was evaluated using the Cognitive Skills Screening Instrument (CASI-S), considering cognitive deficit when scores < 23 in participants aged 60-69 and < 20 in those aged ≥70 years. Foods reported in 24-hour food recall were classified according to their processing level into four groups of NOVA proposal: 1) unprocessed/ minimally processed foods, 2) culinary ingredients, 3) processed foods (products made only from groups 1 and 2); and 4) ultra-processed foods. We estimated the means of total CASI-S score and its four domains according to the quartiles of intake of each food group, and evaluated the association between cognitive decline and each food group intake using logistic models adjusted for gender, age, schooling. Individuals in the highest quartile of unprocessed/minimally processed foods intake had higher scores in temporal orientation (p=0.034), verbal fluency (p=0.002), and total CASI-S score (p=0.004). The scores did differ according to the intake of the other food groups. The ultra-processed was the only group associated with cognitive deficit (OR:1.02; p=0.002). Results suggest nutritional counselling for older adults should focus in reducing ultra-processed and increasing unprocessed foods to help preventing cognitive deficit. Mobile technology offers a remote method to monitor health in older adults and it may provide a platform for early detection of cognitive decline. We aimed to examine attributes of smartphone survey use in the electronic Framingham Heart Study (eFHS) cohort in relation to cognitive testing performed at the time of enrollment. eFHS participants who returned smartphone surveys and underwent cognitive testing